軌跡とシナリオの生成
kinematicTrajectory、waypointTrajectory、および geoTrajectory System object を使用して、センサー モデルを駆動するグラウンド トゥルース軌跡を生成します。trackingScenario を使用して、3 次元アリーナで複数のプラットフォームの追跡をシミュレーションします。トラッキング シナリオ デザイナー アプリを使用して、トラッキング シナリオを設計し、プラットフォームとセンサーを構成し、合成オブジェクト検出を生成します。
関数
アプリ
| トラッキング シナリオ プレーヤー | Play tracking scenario and tracking scenario recording (R2023a 以降) |
| トラッキング シナリオ デザイナー | トラッキング シナリオの設計、プラットフォームとセンサーの構成、合成オブジェクト検出の生成 |
ブロック
| Tracking Scenario Reader | Read tracking scenario and generate simulation data (R2021b 以降) |
| Scenario To Platform | プラットフォーム姿勢をシナリオの座標系からプラットフォームのボディ座標系に変換する (R2022b 以降) |
トピック
- Tracking Simulation Overview
You can build a complete tracking simulation using the functions and objects supplied in this toolbox. The workflow for sensor fusion and tracking simulation consists of three (and optionally four) components. These components are
- Creating a Tracking Scenario
You can define a tracking simulation by using the
trackingScenarioobject. By default, the object creates an empty scenario. You can then populate the scenario with platforms by calling theplatformmethod as many times as needed. A platform is an object (moving or stationary), which can either be a sensor, a target, or any other entity. A platform can be modeled as a point or a cuboid by specifying theDimensionsproperty ofPlatform. After creating a platform, you can specify the motion of the platform by using itsTrajectoryproperty. To configure a trajectory, you can usewaypointTrajectory, which allows you to specify the 3-D waypoints that the platform follows and the associated arrival time for each waypoint. Alternately, you can usekinematicTrajectory, which allows you to specify the 3-D acceleration and angular velocity of the platform with initial pose and translational velocity. You can also specify the orientation of a platform using theOrientationproperty ofkinematicTrajectoryorwaypointTrajectory. - Model Platform Motion Using Trajectory Objects
This topic introduces how to use three different trajectory objects to model platform trajectories, and how to choose between them.
- Define Trajectory Using Positions and Ground Speed
Define
waypointTrajectorywhen time-of-arrival information is not available. (R2023a 以降)
注目の例
Introduction to Tracking Scenario and Simulating Sensor Detections
Introduces how to generate synthetic radar detections in a tracking scenario that simulates target motion and sensor detections. Specifically, this example shows:
Benchmark Trajectories for Multi-Object Tracking
Generate and visualize trajectories of multiple aircraft using trackingScenario and waypointTrajectory.
Multiplatform Radar Detection Generation
Generate radar detections from a multiplatform radar network. The network includes three long-range platforms: two airborne and one ground-based. Such synthetic data can be used to test the performance of tracking architectures for different target types and maneuvers.
Reconstruct Ground Truth Trajectory from Sampled Data Using Filtering, Smoothing, and Interpolation
Use an interactive motion-model based filter along with a smoother to reconstruct the ground truth trajectory based on interpolation techniques.
How to Simulate Out-of-Sequence Measurements
Simulate out-of-sequence measurements (OOSMs). OOSMs occur in multisensor systems when detections from some sensors arrive at the tracker after the tracker has processed detections of the same or earlier time from other sensors. Several reasons, mostly sensor processing lag and sensor-to-tracker connection delays, can cause this problem. For more about OOSMs, see Introduction to Out-of-Sequence Measurement Handling
